Re: DUA and V5UA
"Sergey Mikhailov" <[email protected]>
| Newsgroups | gmane.ietf.sigtran |
|---|---|
| Message-ID | <008001c620e1$7e3e2450$150f0f0a@smikhailov> |
Thank you for the explanation, Mike. Best regards, Sergey Mikhailov. ----- Original Message ----- From: Michael Mentz To: Sergey Mikhailov ; SIGTRAN Sent: Tuesday, January 24, 2006 2:56 PM Subject: RE: [Sigtran] DUA and V5UA Hi Sergey, This is a common question when first looking at the two specifications In essence, the DUA 'extension' should be considered as actually being part of RFC3057 - ie, RFC3057+DUA Extension. All the DUA extension does is introduce some extra messaging and handling in order for the RFC3057 ASP to also support DPNSS interfaces. This follows with the real world use of DPNSS/DASS2, alongside PRI - most vendors offer PRI and DPNSS/DASS2 interfaces on their box - hence the spec takes this into account. Therefore, your RFC3057 ASP has interface identifiers that either use QPTM messaging for PRI/BRI/QSIG, or DPTM messaging for DPNSS/DASS2. For V5UA, the differences are far more fundamental - for one, V5UA uses a different Interface ID format, no doubt this is one of the primary reasons why it is split into it's own protocol. Interfaces Identifiers must be unique within the domain of their ASP, so V5UA cannot run alongside RFC3057 interfaces, and hence must run a separate ASP. Because there is a 1:1 mapping between an ASP and SCTP association, V5UA requires a different SCTP port number. There is probably many other reasons more fundamental than this... As for the SCTP PPID, the only use I have ever seen for this is in tools such as ethereal - it maybe useful to operator to see the QPTM/DPTM data split... Hope this helps, Mike ------------------------------------------------------------------------------ From: [email protected] [mailto:[email protected]] On Behalf Of Sergey Mikhailov Sent: 24 January 2006 07:05 To: SIGTRAN Subject: [Sigtran] DUA and V5UA Hi, I have a question on terminology. DUA (RFC 4129) is considered just as extensions to IUA protocol (even when having its own SCTP Payload protocol ID, no different port number, and some differences from IUA, e.g. in DLCI field format in message header), while V5UA (RFC 3807) is considered a protocol. Both RFCs say that the respective UA "... builds on top of IUA, defining the necessary extensions to IUA for a ... implementation." Is there a valid reason for such differentiation between "extensions to the protocol" and "protocol" in this particular case of DUA and V5UA? Regards, Sergey Mikhailov. _______________________________________________ Sigtran mailing list [email protected] https://www1.ietf.org/mailman/listinfo/sigtran